AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

Self-Supervised Latent Space Optimization With Nebula Variational Coding

TPAMI 2022 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Hierarchical VAEs and posterior collapse · 2 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add the mass term M_a to the nebula loss (vs Euclidean distance to anchor) (ablation)NVC on ForkNet · ShapeNet real completionIoU (%): 24.9 -> 22.1 without M_aImproved
add the mass term M_a to the nebula loss (ablation)NVC on ForkNet · CompleteScanNetIoU (%): 14.1 -> 11.5 without M_aImproved
add L_pair (Siamese) to the metric learning (ablation)NVC-ML on ForkNet · ShapeNet real completionIoU (%): 26.3 -> 24.5 without L_pairImproved
add L_triplet to the metric learning (ablation)NVC-ML on ForkNet · ShapeNet real completionIoU (%): 26.3 -> 23.9 without L_tripletImproved
add L_triplet to the metric learning (ablation)NVC-ML on ScanComplete · CompleteScanNetIoU (%): 19.8 -> 17.3 without L_tripletImproved
add self-supervised metric learning on top of NVC3D-RecGAN · ShapeNet real completionIoU (%): 19.6 -> 20.5Improved
replace GMVAE clustering with NVCNMT-GNMT · WMT13 De-EnBLEU: 29.8 -> 31.2Improved
add the mass term M_a with many anchors (ablation)NVC with 3D-RecGAN · ShapeNet completion, 40 anchorsIoU (%): 74.0 -> 79.3Improved
add NVC to VO-Hand on unoccluded dataVO-Hand (cluster) · Stereo hand poseAUC: 0.984 -> 0.986No clear effect
add NVC to VO-Hand (triplet) on occluded dataVO-Hand · HOP hand poseAUC: 0.597 -> 0.623Improved
add nebula anchors to the latent spacePlanarRecon · NYUv2rel error: 0.134 -> 0.126Improved
add 10 anchors with metric learningNVC-ML on MNIST autoencoder · MNIST reconstructionrel error: 0.169 -> 0.133Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy